The Transition-Season Fault: Neither Hot Nor Cold Decision Tree

Why this matters

Spring and fall calls are the hardest kind of "it is not working right" because nothing is under real load. A heating system is not fighting deep cold, a cooling system is not fighting real heat, and a water system is not stressed by irrigation season yet. Symptoms that would be obvious under full load show up as vague complaints: short cycling, odd noises, a system that "kind of works," a fixture that only misbehaves some mornings. Miss the shoulder-season pattern and you either find nothing (because the fault will not show itself at 65 degrees) or you chase the wrong cause. Read the pattern right and you catch a fault before the first real cold snap or heat wave turns it into an emergency call.

Start here: what shoulder season actually changes

Three things are different in spring and fall versus the peak season, and each one hides or creates different faults:

• Low, inconsistent load. Equipment cycles briefly and often instead of running long steady cycles. Short cycling can be normal shoulder-season behavior or a real fault, and the two look identical from outside.
• Wide daily swings. A 40-degree morning and a 75-degree afternoon put more stress on control logic and materials (expansion and contraction) than a steady extreme does.
• Mode changeover. Systems that switch between heating and cooling, or that get turned on after a season of disuse, are exercising parts and controls that have not moved in months. See the related article on idle equipment for what months of no use does on its own.

Before you diagnose the complaint, ask which of these three is actually driving it. That single question sorts most shoulder-season calls.

If the complaint is short cycling or "won't settle"

1. Check the setpoint against actual conditions first. A system asked to hold a target that is very close to the current ambient will cycle on and off quickly because it satisfies the call almost immediately. This is normal behavior, not a fault, and is the single most common shoulder-season "it keeps turning on and off" call.
2. If cycling happens even with a reasonable setpoint gap, check the control's cycle-rate or minimum-run-time setting. Many controls that ran fine all summer or winter have a protection setting that behaves differently near the changeover point.
3. If cycling is rapid and load is not the explanation, move to a component-level check: a sensor reading erratically, a switch with marginal contacts, or a safety device that is right at its trip threshold under these specific temperatures (a threshold that never got tested at peak load because peak load was months away).
4. If a mechanical part just started moving again after a season of sitting still, expect a brief break-in period. A belt, valve, or damper that sticks for the first few cycles and then runs smoothly is consistent with the idle-equipment pattern, not a new defect. If it does not settle within the first few cycles, treat it as a real fault.

If the complaint is a noise that wasn't there before

Shoulder-season noises are usually thermal-expansion noises, not the mechanical failure they sound like.

• A tick, pop, or creak in the first minutes of a cycle, especially the first cold morning or first hot afternoon of the season, is often ductwork, piping, or a housing expanding or contracting across a wider temperature swing than the system has seen in months. If the noise stops once the system reaches steady temperature and does not recur mid-cycle, this is very likely benign, but confirm there is no rubbing or contact point causing wear, since a bracket or clamp that has loosened over the season can turn a benign expansion tick into real chafing damage over time.
• A grinding, squealing, or knocking noise that continues through the cycle, not just at startup, points to a mechanical issue (bearing, belt, fan) that idle time or the seasonal changeover exposed. Treat this as a real fault, not a transition noise, and inspect the moving component directly.
• A noise only on mode changeover (switching between heating and cooling, or a valve reversing) that clears within a few seconds is usually the changeover mechanism doing its job. A noise that persists after the changeover completes is not.

If the complaint is "it works sometimes"

This is the hardest shoulder-season pattern because it is genuinely intermittent, driven by the daily temperature swing rather than a constant condition.

1. Ask what time of day it fails. A fault that only shows up in the cold early morning or the hot late afternoon is temperature-triggered: a sensor, a control board, or a mechanical tolerance that behaves differently at the extremes of the daily swing even though the swing itself is mild by peak-season standards.
2. If it fails only at the coldest or hottest point of the day, that is your reproducible window. Schedule the visit to be on site during that window rather than mid-day when nothing will show.
3. If it fails at random times with no clear temperature link, widen the suspect list beyond seasonal cause: a loose connection, a marginal component, or a control fault that shoulder-season cycling (more starts and stops per day than peak season) is simply exercising more often and therefore exposing sooner.
4. If the system was off or unused for the off-season, run the idle-equipment checklist (fluid condition, backup battery, pest intrusion, control settings and clock) before assuming a new defect. A large share of "it worked fine last time we used it" complaints trace back to months of sitting idle, not a fault that developed during use.

If the complaint follows the first real cold snap or heat wave after a mild stretch

This is a different pattern from the above: a system that ran fine through the mild shoulder-season swings and then fails the moment real load arrives.

1. Suspect a marginal component that mild conditions were not stressing. A capacitor, a weak connection, a slightly undersized part, or accumulated debris will often pass at low load and fail at full load.
2. Do not dismiss it as coincidence. A failure that lines up exactly with the first hard load of the season almost always is load-related, not a random new fault.
3. Load-test deliberately if you can. Where safe and practical, running the system at a higher demand than the shoulder season provided will reproduce the fault faster than waiting for the next real event.

Recap

1. Sort the complaint into cycling, noise, intermittent, or first-real-load-failure.
2. Check setpoint-versus-ambient and idle history before assuming a component fault.
3. Match noises to their timing (startup-only vs continuous, changeover-only vs persistent).
4. For intermittent faults, find the reproducible window (time of day, temperature extreme) instead of guessing.
5. Treat a fault that appears exactly at the season's first real load as load-related, not coincidental.

References

• Manufacturer documentation on seasonal startup and changeover procedures
• Trade-standard practice for shoulder-season commissioning checks
• See related: What Sitting Idle for Months Does to Equipment
• See related: Reading a Fleet-Wide Failure Pattern